WEBVTT 1 00:00:00.841 --> 00:00:11.330 Hey everyone, welcome to Future Ready Today, the number one podcast focused on the future of work, where every day I break down some of the top stories that you need to be paying attention to that are shaping the future of work. 2 00:00:11.370 --> 00:00:19.638 I summarize them for you and I give you the signals and the trends that you need to be paying attention to so that you can, of course, be future ready today. 3 00:00:20.339 --> 00:00:22.801 It is Tuesday, March 10th, 2026. 4 00:00:22.841 --> 00:00:24.742 There are five stories that I want to get to today. 5 00:00:25.623 --> 00:00:33.393 One is going to be from Fortune, talking about how AI just handed workers back six extra hours and their bosses immediately took it away. 6 00:00:34.194 --> 00:00:37.458 Story number two, highlighting a KPMG story. 7 00:00:37.858 --> 00:00:44.326 A survey of 100 major CEOs find that only 9% plan to cut jobs this year because of AI. 8 00:00:45.913 --> 00:00:54.323 A third story, China launched the most ambitious society-wide AI employment push in history. 9 00:00:55.084 --> 00:00:57.046 Story number four from Fortune. 10 00:00:57.867 --> 00:01:04.635 They have an important piece that the one number that CEOs are now using to quietly recalculate 11 00:01:05.442 --> 00:01:07.123 how many humans they actually need. 12 00:01:07.843 --> 00:01:19.329 And the last story is The Verge went deep on Merkur, the $10 billion startup paying white collar professionals to train the AI that may eventually replace them. 13 00:01:20.250 --> 00:01:26.213 So the five stories we're gonna get to today before I jump into those five stories. 14 00:01:26.353 --> 00:01:28.614 I have a post going up later today 15 00:01:29.214 --> 00:01:36.838 This week on my Substack, it's going to be specifically talking about how AI is impacting HR. 16 00:01:37.578 --> 00:01:48.183 So if you're an HR professional, you're a CHRO, you're an entry-level HR leader, you are definitely going to want to make sure that you subscribe to the Substack and get access to this. 17 00:01:49.143 --> 00:01:52.185 You can go to futureofworknewsletter.com. 18 00:01:53.821 --> 00:01:58.004 Again, that is futureofworknewsletter.com. 19 00:01:58.664 --> 00:02:03.787 The last few posts that I put up there have garnered a lot of attention, a lot of discussion. 20 00:02:03.807 --> 00:02:05.508 I got quite a few emails from people. 21 00:02:06.569 --> 00:02:11.111 One, the most recent one was 17 charts that reveal the astonishing scale of AI. 22 00:02:12.232 --> 00:02:13.614 and technological change. 23 00:02:14.215 --> 00:02:17.901 So again, to get access to that, go to futureofworknewsletter.com. 24 00:02:18.422 --> 00:02:21.086 There is a paid version, which is $9.99 a month. 25 00:02:21.226 --> 00:02:23.410 That's how you get access to these deep dive articles. 26 00:02:24.011 --> 00:02:25.072 There's also a free version. 27 00:02:26.093 --> 00:02:30.054 In the free version, you get access to my weekly briefings. 28 00:02:30.474 --> 00:02:32.535 So I'll talk about some of the top stories of the week. 29 00:02:33.435 --> 00:02:38.836 I'll explain to you the significance of those stories and what they mean for you. 30 00:02:39.196 --> 00:02:40.076 That is totally free. 31 00:02:40.156 --> 00:02:43.037 All you got to do is enter in your email to get access to those. 32 00:02:43.757 --> 00:02:47.798 Again, that's all at futureofworknewsletter.com. 33 00:02:48.518 --> 00:02:51.399 All right, let's jump right into the stories of the day. 34 00:02:51.419 --> 00:02:55.540 The first one, this one published by Fortune, which by the way, I got a compliment. 35 00:02:56.462 --> 00:03:00.965 They've been doing a very good job lately covering some interesting future work stories. 36 00:03:01.945 --> 00:03:07.829 So the title of this one, AI just gave you six extra hours back and your boss already took them. 37 00:03:08.469 --> 00:03:16.454 This post was written by Nick Lichtenberg and he opens up with a line that I think captures an interesting moment. 38 00:03:17.314 --> 00:03:21.757 And he says, tasks that once took six hours now take less than one. 39 00:03:22.878 --> 00:03:28.987 A two-week process can sometimes be finished in an afternoon, but workers are not getting their time back. 40 00:03:29.748 --> 00:03:32.792 So Fortune spoke to Google Cloud's Yasmin Ahmad. 41 00:03:32.892 --> 00:03:39.481 She is, I'm sorry, who's the executive that Fortune 500 companies call when they want to figure out how to put AI work at scale. 42 00:03:40.602 --> 00:03:48.304 She told Fortune that efficiency gains are very real and that executives are just trying to keep them quiet. 43 00:03:49.104 --> 00:03:51.465 She said organizations are, in her words, nervous. 44 00:03:51.905 --> 00:03:59.287 But in private conversations, she said executives admit they are thinking hard about what all of these efficiencies really mean. 45 00:04:00.087 --> 00:04:05.889 So there are a couple data points that were highlighted in this article, one from AES Energy. 46 00:04:07.046 --> 00:04:15.749 A major energy company, they transformed a 14-day auditing and data entry process into a task that now takes one hour. 47 00:04:16.670 --> 00:04:18.030 14 days to one hour. 48 00:04:18.910 --> 00:04:22.052 Dun & Bradstreet, they are the data and analytics giant. 49 00:04:22.112 --> 00:04:22.832 They shrink. 50 00:04:23.718 --> 00:04:24.378 Number crunching. 51 00:04:25.079 --> 00:04:26.340 From hours to minutes. 52 00:04:27.300 --> 00:04:30.762 Dun & Bradstreet's CTO Mike Manos put it plainly. 53 00:04:30.802 --> 00:04:34.904 He said, I got the eight hours down to two hours. 54 00:04:35.505 --> 00:04:39.387 But now I can get 20 hours of work because the work came down. 55 00:04:40.952 --> 00:04:49.615 Product development cycle that was tracking to take 24 to 36 months, got done in six months, and nobody got sent home early. 56 00:04:50.095 --> 00:04:58.358 At Google, 50% of all code is now being written by AI and producing over a 10% velocity gain across tens of thousands of engineers. 57 00:04:58.978 --> 00:05:11.965 KPMG CEO Tim Walsh told Fortune that time spent preparing for executive meetings dropped by around 75% after deploying Gemini, and within two weeks of launch, over 90% of KPMG professionals were using it. 58 00:05:13.038 --> 00:05:16.542 And his response was, my business should be growing and it will grow. 59 00:05:17.143 --> 00:05:20.227 I see the number of my employees going up, not down. 60 00:05:20.928 --> 00:05:31.901 But Wharton professor Peter Capelli, who's been on this podcast, I think once or twice, he pushed back and he pointed to a Boston consulting group study that coined the term AI brain fry. 61 00:05:32.642 --> 00:05:44.691 saying that workers who constantly supervise multiple AI tools report 12% more mental fatigue and significantly more information overload and decision fatigue. 62 00:05:45.472 --> 00:05:47.573 So some interesting things here to be paying attention to. 63 00:05:47.593 --> 00:05:53.618 The article also draws a direct line to John Maynard Keyes, who predicted in the 1930s that by the year 2030, 64 00:05:55.836 --> 00:05:58.879 a 15-hour workweek would be possible. 65 00:05:59.900 --> 00:06:03.363 And he then famously worried, what is everybody going to do with their free time? 66 00:06:04.464 --> 00:06:09.389 And so the interesting thing here is, is that actually going to happen or not? 67 00:06:09.689 --> 00:06:17.637 I highlighted Vinod Khosla's statements, I think it was last week, where he said that five-year-olds today are not even going to need to have jobs. 68 00:06:18.642 --> 00:06:23.626 So the futurist lens, and that is the question that gets asked frequently. 69 00:06:24.006 --> 00:06:28.469 I've addressed it last week as well, and that is who captures the gains? 70 00:06:29.130 --> 00:06:34.454 So if AI is going to have this impact across productivity gains, who captures it? 71 00:06:35.474 --> 00:06:39.878 And right now, the answer is not the workers, not the workers. 72 00:06:40.987 --> 00:06:49.313 So since 1979, the United States has grown roughly 65%, meaning productivity has grown by 65%. 73 00:06:50.394 --> 00:06:54.717 Typical worker compensation over that same time period has grown by about 18%. 74 00:06:55.017 --> 00:07:00.882 And AI is not creating that gap, it's actually turbocharging, supercharging that gap. 75 00:07:01.662 --> 00:07:03.203 But the mechanism now is a little bit different. 76 00:07:03.404 --> 00:07:04.624 So in the industrial era, 77 00:07:05.445 --> 00:07:08.246 Productivity gains from machinery took decades to show up. 78 00:07:08.346 --> 00:07:14.268 And as I've talked about many times on this podcast, now the gains are pretty much immediate. 79 00:07:14.588 --> 00:07:16.129 So we're not going from decades anymore. 80 00:07:16.149 --> 00:07:20.070 We're going to a span of days, weeks, months at the most. 81 00:07:21.090 --> 00:07:24.791 And this is what makes this time categorically different. 82 00:07:26.392 --> 00:07:29.653 On top of that, the AI brainfire research reveals something interesting. 83 00:07:30.706 --> 00:07:31.447 Interesting as well. 84 00:07:32.509 --> 00:07:36.114 And that's something a lot of the productivity people in this space are missing. 85 00:07:36.654 --> 00:07:39.398 And that is that cognitive load is not linear. 86 00:07:39.538 --> 00:07:45.827 We've been measuring productivity in outputs, but human performance runs on a non-linear curve. 87 00:07:46.508 --> 00:07:49.431 So moderate challenge sharpens performance. 88 00:07:49.891 --> 00:07:55.015 Chronic overload degrades memory, judgment, and emotional regulation. 89 00:07:55.696 --> 00:08:01.340 And these are precisely the capabilities that differentiate the human workers from their AI counterparts. 90 00:08:02.121 --> 00:08:09.443 And so companies are borrowing against biological reserves and calling it productivity, calling it a productivity win. 91 00:08:10.143 --> 00:08:22.927 But the debt gets paid later in burnout, in disengagement, in turnover, and deteriorating quality of the very judgment-intensive work that AI supposedly can't do. 92 00:08:24.067 --> 00:08:27.628 So again, a little bit of a double-edged sword here. 93 00:08:28.663 --> 00:08:31.644 There's also a little bit of a macro issue that people aren't talking about. 94 00:08:32.164 --> 00:08:39.728 You cannot extract infinite output from a shrinking labor share and still grow a consumer company. 95 00:08:40.248 --> 00:08:42.669 Henry Ford understood this in 1914. 96 00:08:43.109 --> 00:08:49.832 He doubled wages, not because he was trying to be charitable, but because he needed workers who could afford to buy his cars. 97 00:08:50.492 --> 00:08:59.396 So when consumers fuel 70% of the economy, as Fortune's Diane Brady pointed out elsewhere today, you need people who have money to spend. 98 00:09:00.216 --> 00:09:11.942 And so AI that generates productivity for companies but remains stagnant or declines real incomes for workers is a growth engine that eventually runs out of fuel. 99 00:09:12.442 --> 00:09:19.625 Because if workers aren't growing, they're not making more money, who cares about their productivity gains if you don't have people who are buying those products and services? 100 00:09:20.658 --> 00:09:26.221 Story two, more CEOs envision hiring than firing due to AI. 101 00:09:27.401 --> 00:09:29.042 Story published by Axios. 102 00:09:29.682 --> 00:09:39.427 They're reporting today on a new KPMG US CEO Outlook Pulse survey from 100 CEOs at companies with revenues over $500 million. 103 00:09:40.107 --> 00:09:43.389 They were surveyed between January and mid-February. 104 00:09:44.239 --> 00:09:52.062 And the headline finding, only 9% of those CEOs plan to reduce their workforce because of AI in 2026. 105 00:09:52.182 --> 00:10:00.485 55% expect to increase hiring as a direct result of AI, and 36% expect no change. 106 00:10:01.265 --> 00:10:03.986 Now, that is, of course, the optimistic version. 107 00:10:04.746 --> 00:10:07.067 I think there is a more important version of 108 00:10:09.295 --> 00:10:09.915 Sorry, excuse me. 109 00:10:09.935 --> 00:10:15.917 My whole house had a terrible cough, so I'm powering through on hot tea and cough drops these days. 110 00:10:16.698 --> 00:10:28.342 KPMG CEO Tim Walsh told Axios that the majority of companies right now are not actually realizing, nor can they see, the return on investment of the AI they're deploying. 111 00:10:29.042 --> 00:10:37.577 In other words, they're spending massively, they're bullish on the long term, and most of them cannot yet point to what they're actually getting for it. 112 00:10:37.657 --> 00:10:42.245 I found this the same from all the CHOs that I interview who are part of my group as well. 113 00:10:43.802 --> 00:10:48.463 9 in 10 CEOs are now worried about malware and phishing attacks powered by AI. 114 00:10:48.883 --> 00:10:53.704 Nearly 6 in 10 are worried about quantum computing attacks on encrypted data. 115 00:10:54.344 --> 00:11:06.787 And 77% surveyed agree that generative AI was overhyped over the past year, but also that its disruptive potential over the next 5 to 10 years is likely to be underhyped. 116 00:11:07.700 --> 00:11:18.204 Now, the futurist lens here, the 9% and the 55% numbers are going to be cited everywhere, I think, as evidence that the AI jobs apocalypse is not happening. 117 00:11:19.124 --> 00:11:27.447 And in a narrow technical sense for this particular group for 2026, that may be accurate, but there are a few things that we should examine underneath. 118 00:11:28.427 --> 00:11:30.808 So, first of all, hiring... 119 00:11:32.146 --> 00:11:37.229 hiring more is not the same as displacement not happening. 120 00:11:38.149 --> 00:11:45.092 Companies can grow headcount in aggregate while simultaneously eliminating entire job categories. 121 00:11:45.693 --> 00:11:56.738 You can hire 50 new AI orchestration engineers and let 200 junior analysts disappear through attrition and never trigger a mass layoff event. 122 00:11:57.579 --> 00:11:59.720 So the net number looks positive, 123 00:12:00.428 --> 00:12:04.810 But the lived reality for those 200 workers, of course, is completely invisible in the data. 124 00:12:05.991 --> 00:12:15.536 Second, the ROI admission is the most important thing in this story, and it's been common over the past, I don't know, three months since organizations have been talking about this. 125 00:12:16.376 --> 00:12:23.940 Gartner projects $2.5 trillion in global AI capital spending this year, which is staggering. 126 00:12:25.518 --> 00:12:28.820 but most CEOs yet cannot see the return of that investment. 127 00:12:29.581 --> 00:12:36.725 That means we are investing very heavily and in the pre-productivity phase of the curve. 128 00:12:37.125 --> 00:12:42.589 And this is the valley where a lot of capital gets deployed, but the gains have not yet materialized. 129 00:12:42.649 --> 00:12:54.637 And history shows that the productivity gains do eventually come, but the lag between the investment and the realized productivity is exactly when the workforce disruption actually accelerates. 130 00:12:55.177 --> 00:13:05.628 Because companies are under enormous pressure during this period between point A and point B to cut costs and to justify the capital that they're putting into these AI tools. 131 00:13:06.649 --> 00:13:08.792 And third, the cybersecurity number. 132 00:13:10.008 --> 00:13:17.314 is I think stunning and underreported, and that is that nine out of 10 CEOs worried or are worried about AI-powered cyber attacks. 133 00:13:18.035 --> 00:13:19.876 That is not a fringe concern. 134 00:13:19.916 --> 00:13:27.142 It means the same technology being used to democratize productivity is also democratizing the ability to attack. 135 00:13:28.082 --> 00:13:38.509 And the chief information security officer and the people function of organizations are going to have to converge in ways that most companies have not yet started planning for. 136 00:13:39.250 --> 00:13:49.257 So in general, looking at this, it seems like CEOs are optimistic about AI's long-term potential, but they are quite cautious about the near-term returns. 137 00:13:49.697 --> 00:13:53.820 They're not pulling the mass layoff trigger yet. 138 00:13:53.900 --> 00:13:55.141 And hopefully many of them won't. 139 00:13:56.001 --> 00:14:06.444 And they're almost certainly underestimating, I think, how quickly the calculus of all of this is going to change once the technology crosses the next capability threshold. 140 00:14:06.904 --> 00:14:11.365 So, for example, we're at ChatGPT 5.4, just got released last week. 141 00:14:12.565 --> 00:14:13.926 What do we think is going to happen next? 142 00:14:15.312 --> 00:14:21.278 In the next three to six months, when we're on ChatGPT 5.7 or 5.8 or Claude Opus 4.9 or 5.0 or Gemini 3.6. 143 00:14:29.630 --> 00:14:32.612 These models are improving rapidly. 144 00:14:32.833 --> 00:14:37.716 So by the time we get to that point, we might be having a very different conversation here. 145 00:14:37.736 --> 00:14:49.024 And we're just not sure, at least us on the outside, are just not sure how quickly these AI tools are actually evolving, what is next for subsequent versions. 146 00:14:49.745 --> 00:14:52.527 We don't have that insight, so it becomes very hard to plan. 147 00:14:53.187 --> 00:14:54.789 I mean, there was a pretty big jump. 148 00:14:54.809 --> 00:14:59.635 I think it was between ChatGPT 2.something and 3.0. 149 00:14:59.915 --> 00:15:02.638 It seemed like an exponential growth jump. 150 00:15:03.580 --> 00:15:09.887 What if we have something like that again between ChatGPT 5.4 and 6.0? 151 00:15:10.868 --> 00:15:12.110 So it remains to be seen. 152 00:15:14.294 --> 00:15:22.537 Story three, and I think this is a fairly interesting one as well, China pins hopes on society-wide AI push to add jobs and rejuvenate their economy. 153 00:15:22.617 --> 00:15:23.738 This is a Reuters story. 154 00:15:24.518 --> 00:15:36.323 Reuters is reporting today from Beijing on China's most ambitious AI employment strategy to date, unveiled at the annual session of the National People's Congress. 155 00:15:36.943 --> 00:15:50.629 China plans to go all in on AI as a job creation engine over the next five years, explicitly framing it as the tool for offsetting an aging workforce and long-term economic slowdown. 156 00:15:51.429 --> 00:15:53.810 So here's the scale of what we're talking about here. 157 00:15:54.150 --> 00:16:00.213 There are about 300 million Chinese workers that are expected to retire in the next decade. 158 00:16:00.933 --> 00:16:05.718 That is almost the entire working age population of the United States. 159 00:16:06.439 --> 00:16:10.984 China's new five-year plan mentions AI more than 50 times and includes... 160 00:16:12.229 --> 00:16:18.895 a sweeping AI plus action plan to integrate the technology across every major industry. 161 00:16:19.615 --> 00:16:26.501 Policymakers and company executives are actively downplaying global fears that AI will stunt unemployment. 162 00:16:27.362 --> 00:16:39.973 Analyst Xu Jing-hi of consultancy Plenum told Reuters, For now, advancing AI adoption and capability appears to be a higher policy priority than preemptively addressing potential job displacement. 163 00:16:40.613 --> 00:16:48.872 She also noted that the emphasis on job creation leaves Beijing room to respond if more disruptive labor market effects become evident. 164 00:16:49.931 --> 00:16:53.194 Universities are already reworking their curriculums. 165 00:16:53.714 --> 00:17:02.202 Shanghai Tech University, for example, has introduced AI micro-majors specifically focused on skills AI cannot easily replace. 166 00:17:02.763 --> 00:17:06.386 Cross-disciplinary learning, critical thinking, and creativity. 167 00:17:06.926 --> 00:17:09.689 The provost's statement to Reuters is worth quoting directly. 168 00:17:09.709 --> 00:17:11.110 The provost's statement to Reuters is worth quoting directly. 169 00:17:11.250 --> 00:17:14.071 He says, we must train them to ask questions. 170 00:17:14.531 --> 00:17:18.393 If your thinking isn't sharp, you won't beat the robots. 171 00:17:19.113 --> 00:17:30.257 China's five-year plan also set a lower growth target of 4.5 to 5% for 2026 in terms of GDP growth, down from last year's 5%. 172 00:17:30.517 --> 00:17:33.718 And I think this is the lowest over the last many, many years. 173 00:17:35.078 --> 00:17:38.900 And this is as Beijing works to rebalance an economy that invests 20%. 174 00:17:40.378 --> 00:17:49.214 percentage points of GDP more than the global average while its households spend roughly 20 points less. 175 00:17:50.356 --> 00:17:51.458 So there is, 176 00:17:53.698 --> 00:17:55.020 an interesting thing happening in China. 177 00:17:55.661 --> 00:18:04.693 And say what you will about China, but they are taking AI seriously and it's being pushed and mandated in a lot of these schools. 178 00:18:05.354 --> 00:18:10.080 I wish our educational institutions in the United States would be having more conversations like this. 179 00:18:11.070 --> 00:18:15.571 Now, the futurist lens here is that China is running what I think is a little bit of a national wager. 180 00:18:16.251 --> 00:18:26.053 And the wager that they are betting on is that AI is actually going to create more jobs than it destroys, fast enough to absorb a demographic cliff of historic proportions. 181 00:18:26.133 --> 00:18:32.355 Now, they're obviously a unique percentage or unique situation because of the high percentage of employees who are going to be retiring. 182 00:18:33.844 --> 00:18:39.786 It's a pretty audacious bet, I think, but it makes complete sense when you understand China's structural problem. 183 00:18:40.386 --> 00:18:50.970 They have a shrinking working age population, slowing productivity growth, an economic model that has historically depended on a massive human labor inputs. 184 00:18:51.771 --> 00:18:55.452 And in that context, I think it's safe to say that AI adoption is not a luxury. 185 00:18:56.012 --> 00:18:57.633 It's not some sort of a competitive play. 186 00:18:58.133 --> 00:19:00.294 It is a survival mechanism for them. 187 00:19:00.314 --> 00:19:00.854 They have to do it. 188 00:19:02.404 --> 00:19:09.328 The policy framing of AI as job creating rather than job displacing is, you can make the argument. 189 00:19:09.408 --> 00:19:15.111 It's partially ideological, but I think it's also very, very strategic. 190 00:19:15.871 --> 00:19:27.217 A government that tells 1.4 billion people that AI will take their jobs, well, you can imagine what kind of social instability and chaos that is going to cause. 191 00:19:28.257 --> 00:19:34.841 And so a government that reframes AI as a rejuvenation, they get a very different compliance curve from its population. 192 00:19:35.762 --> 00:19:43.006 Now, the Shanghai Tech Micro Major Model, I think is one of the most interesting workforce experiments happening anywhere in the world today. 193 00:19:43.727 --> 00:19:53.533 The explicit focus on capabilities that AI cannot replace and cannot take away from people, things like asking good questions, reasoning across disciplines, being creative. 194 00:19:54.854 --> 00:19:57.576 I think this is a direct acknowledgement 195 00:19:58.260 --> 00:20:01.963 that the commodity skill playbook of the past three decades is over. 196 00:20:01.983 --> 00:20:09.529 And as I mentioned, I wish companies in the States and educational institutions in the States would be having these types of conversations as well on a national level. 197 00:20:10.289 --> 00:20:12.031 And the insight is not uniquely Chinese. 198 00:20:12.071 --> 00:20:23.140 I think, as I mentioned, it applies to every education system, every training program, every individual trying to figure out where to invest their human capital over the next 10 years. 199 00:20:23.880 --> 00:20:25.521 I think there's also a little bit of a geopolitical, 200 00:20:27.022 --> 00:20:30.944 you can call it a tension or dimension that we should be paying attention to. 201 00:20:31.564 --> 00:20:41.848 The United States and China are both running AI workforce strategies right now, and we're competing very heavily against each other, but we're running at very different scales of coordination. 202 00:20:42.508 --> 00:20:50.731 China's five-year plan aligns universities, companies, government agencies, monetary policy around a single technology strategy. 203 00:20:50.751 --> 00:20:54.913 In other words, their entire economy is focused on this. 204 00:20:55.993 --> 00:20:58.435 The United States has market innovation. 205 00:20:58.735 --> 00:21:00.557 We have entrepreneurial velocity. 206 00:21:01.097 --> 00:21:16.209 We have individual company initiatives, but nothing approaching a coordinated national response when it comes to AI that is touching government and policy and companies and education and monetary and financial policy, social safety. 207 00:21:16.729 --> 00:21:20.011 We're just not thinking about it from a cohesive standpoint the way that China is. 208 00:21:20.752 --> 00:21:26.913 And I think the asymmetry can potentially have consequences that most organizations are just not factoring into their planning. 209 00:21:28.453 --> 00:21:29.614 Story four from Fortune. 210 00:21:29.934 --> 00:21:31.614 Again, Fortune, man, they're on fire today. 211 00:21:31.694 --> 00:21:36.595 CEOs are using one number in the AI age to decide how many people they still need. 212 00:21:37.695 --> 00:21:39.895 I think this is one of the more interesting stories published today. 213 00:21:40.856 --> 00:21:48.577 And it's built around a single metric that KPMG CEO Tim Walsh says is quietly reshaping how corporate America is thinking about headcount. 214 00:21:49.318 --> 00:21:54.861 And it's not revenue per employee, which has anchored workforce decisions for many decades. 215 00:21:55.561 --> 00:21:57.863 It's not productivity in the traditional sense. 216 00:21:58.563 --> 00:22:01.084 Walsh calls it labor costs margin. 217 00:22:01.605 --> 00:22:12.270 And once you understand it, it can potentially reveal more about where AI is actually taking the economy than almost anything being said out loud in boardrooms and executive sessions. 218 00:22:12.911 --> 00:22:13.651 So here's how it works. 219 00:22:15.218 --> 00:22:25.405 For every business engagement, CEOs are now asking, what is my mix of labor versus technology, and what is the cost of actually delivering that work? 220 00:22:26.365 --> 00:22:29.007 The direction of travel is unambiguous. 221 00:22:29.847 --> 00:22:39.854 Lower labor cost per unit of work, higher technology cost, more total volume pushed through the business at a lower cost per unit. 222 00:22:40.598 --> 00:22:46.860 The net result, companies can scale output without proportionally scaling headcount. 223 00:22:47.881 --> 00:22:50.201 So let me repeat that really quick. 224 00:22:51.302 --> 00:22:53.322 Lower labor cost per unit of work. 225 00:22:54.543 --> 00:22:55.783 Higher technology cost. 226 00:22:57.004 --> 00:23:02.386 More total volume getting pushed through the business at a lower cost per unit. 227 00:23:02.486 --> 00:23:03.126 So in other words... 228 00:23:03.915 --> 00:23:17.506 you're investing less in labor, more in technology, and this allows you to push more work, more volume through the business, and the result is that you get a lower cost per unit. 229 00:23:18.387 --> 00:23:30.977 And what this means, as I just mentioned, you can scale the output of the products or services that you produce without proportionally scaling headcount with that same ratio. 230 00:23:31.833 --> 00:23:43.724 In the KPMG survey of 100 CEOs that I mentioned, 77% believe generative AI was overhyped in the past year, but also that its disruptive potential over the next 5 to 10 years is underhyped. 231 00:23:44.605 --> 00:23:52.092 Nearly 80% are allocating at least 5% of their total capital budgets to AI. 232 00:23:52.132 --> 00:23:53.873 41% are putting in at least 10%. 233 00:23:55.192 --> 00:24:02.641 and 35% are spending between 11 and 20% of their entire capital budget on technology. 234 00:24:03.361 --> 00:24:04.202 Now for context, 235 00:24:04.945 --> 00:24:15.910 That level of capital allocation rivals what companies were devoting to cloud infrastructure at the height of the cloud transition, and cloud took a decade to fully reshape the economy. 236 00:24:16.731 --> 00:24:20.753 And so Walsh was clear about which jobs are in the most danger. 237 00:24:20.813 --> 00:24:31.618 He said, quote, You can look at those types of jobs that are repetitive tasks, people that are doing the same thing every single day in and day out, and that's a scary place to be right now. 238 00:24:32.218 --> 00:24:37.041 His own hiring at KPMG has not gone down, but the composition has shifted fundamentally. 239 00:24:37.501 --> 00:24:46.767 He's now hiring orchestrators, people managing large AI-driven workloads to make sure the outputs are complete, accurate, and are reaching the right destination. 240 00:24:48.168 --> 00:24:59.335 Interesting that he mentions orchestrators because in my new book, The Eight Laws of Employee Experience, I talk about five new archetypes of leaders, 241 00:25:00.398 --> 00:25:03.320 One of those archetypes is the orchestrator. 242 00:25:04.141 --> 00:25:12.147 And so if you're interested in learning more about what the future of work is gonna look like over the next five to 10 years, you're gonna wanna grab a copy of this book 243 00:25:12.945 --> 00:25:15.346 The number 8exlaws.com. 244 00:25:15.386 --> 00:25:17.527 That's the number 8exlaws.com. 245 00:25:17.547 --> 00:25:23.810 The title of the book, The Eight Laws of Employee Experience, How to Build a Future Ready Organization. 246 00:25:24.371 --> 00:25:28.513 And again, I wrote this long before Tim Walsh has been saying this. 247 00:25:28.593 --> 00:25:31.654 So technically, I'm ahead of the CEO of KPMG, right? 248 00:25:32.335 --> 00:25:36.477 Anyway, if you want to get some of those insights, again, the number 8exlaws.com. 249 00:25:37.960 --> 00:25:46.065 Okay, getting back to the story here, two thirds of CEOs admitted that they have not yet redefined roles or career paths to account for AI. 250 00:25:47.186 --> 00:25:58.513 And 31% said their top concern about AI's impact on leadership development is the reduced opportunities for early career employees to build judgment through real world experience. 251 00:26:00.375 --> 00:26:05.518 So now the futurist lens here, and I just had a conversation with somebody about this. 252 00:26:06.625 --> 00:26:20.060 And I was telling them that I think the most important thing for employees across all levels to do now is to have conversations with your leaders, with your managers, with your peers about your job, your role, and the use of AI. 253 00:26:21.321 --> 00:26:26.526 And sometimes people say, well, don't the organizations have a responsibility now? 254 00:26:27.418 --> 00:26:31.942 to employees to convey to them, what are the roles that are going to be displaced? 255 00:26:32.602 --> 00:26:34.003 What are the skills that are going to be relevant? 256 00:26:34.544 --> 00:26:36.866 How can you still have a role at this company in the next few years? 257 00:26:38.327 --> 00:26:39.468 The answer to that is yes. 258 00:26:40.309 --> 00:26:43.251 But the problem is that companies themselves don't know. 259 00:26:44.552 --> 00:26:46.374 So you're in a very difficult spot. 260 00:26:47.655 --> 00:26:48.876 If you rely on your company... 261 00:26:49.909 --> 00:27:00.975 to tell you what you should do to reskill, to upskill, where the opportunities are, where the growth is, by the time they're even having that conversation with you, it might be too late. 262 00:27:02.176 --> 00:27:06.898 You need to start having the conversations and initiating these things yourself. 263 00:27:07.739 --> 00:27:19.345 So for example, if you're in customer service, you know for a fact, or in finance for example, you know that AI has a lot of potential to disrupt your role and your job. 264 00:27:20.324 --> 00:27:25.927 So you should be doing your own research right now to understand what potential impact AI might have. 265 00:27:26.588 --> 00:27:29.790 Where are the jobs, the industries that might be impacted? 266 00:27:29.930 --> 00:27:32.171 And just go to your leader and say, hey, can we grab coffee? 267 00:27:33.512 --> 00:27:35.113 I'm worried about AI in my role. 268 00:27:35.733 --> 00:27:40.296 If AI did come and automated me, where do you think I could potentially move in the company? 269 00:27:40.837 --> 00:27:42.418 Go talk to those hiring managers. 270 00:27:42.518 --> 00:27:44.139 Say, hey, look, just to be safe. 271 00:27:44.759 --> 00:27:46.100 I want to take some time on my own. 272 00:27:46.120 --> 00:27:50.262 I'm going to try to reskill myself, upskill, learn some new areas in the business. 273 00:27:50.682 --> 00:27:59.207 That way, if AI ever comes from me, I just want you to know that I also have skills where I can move from customer service into product management or into marketing or into sales. 274 00:27:59.947 --> 00:28:01.028 I want to stay at this company. 275 00:28:01.348 --> 00:28:02.188 I'm going to be prepared. 276 00:28:02.849 --> 00:28:08.732 And again, the rationale for that is that companies themselves do not know. 277 00:28:09.572 --> 00:28:10.613 They just don't. 278 00:28:12.195 --> 00:28:15.618 Companies are not doing these skills maps yet. 279 00:28:15.838 --> 00:28:21.283 They're not sure where is AI going to impact the business and how. 280 00:28:21.703 --> 00:28:23.025 So they don't have answers for you. 281 00:28:24.086 --> 00:28:27.869 So if you don't know, and if they don't know, it creates a difficult spot. 282 00:28:27.929 --> 00:28:30.592 So I want you to think of yourself as a detective. 283 00:28:31.633 --> 00:28:40.058 Look for the clues, look for the signs, have the conversations and try to proactively prepare yourself for things that might happen. 284 00:28:41.139 --> 00:28:49.584 Now, the other element of this that came up during my conversation is, are people even aware of what's happening with AI? 285 00:28:50.485 --> 00:28:51.666 I think a lot of people are. 286 00:28:53.006 --> 00:29:00.211 I think some of the recent research that I saw showed that if you're on the front line, you're perhaps less aware than if you're kind of on the knowledge work side. 287 00:29:01.674 --> 00:29:05.237 But it's not that people are not aware of AI. 288 00:29:05.537 --> 00:29:07.239 I don't think they're aware of the time horizon. 289 00:29:08.380 --> 00:29:13.244 So you might be in a role like customer service and thinking like, okay, I know AI is out there. 290 00:29:13.524 --> 00:29:17.308 I know it will eventually impact my role, but I got two, three, four, five years. 291 00:29:18.629 --> 00:29:26.135 And then three months down the road, your company's coming over to you and saying, hey, we're letting go of 1,000 employees in customer service. 292 00:29:26.155 --> 00:29:26.836 And you're like, wait a minute. 293 00:29:27.737 --> 00:29:28.898 I thought I had three to five years. 294 00:29:29.924 --> 00:29:31.604 So there are two things that you need to be aware of. 295 00:29:32.285 --> 00:29:32.985 One is the impact. 296 00:29:33.045 --> 00:29:34.325 The second is the time horizon. 297 00:29:35.025 --> 00:29:40.487 And as I mentioned, companies do not have answers for you on either of these things. 298 00:29:41.547 --> 00:29:52.030 So you as an individual and as an employee have to put more accountability and responsibility into your own due diligence for this. 299 00:29:52.350 --> 00:29:56.891 And I know it's a tough spot to be in, but it's the only spot that you can actually be in. 300 00:29:57.172 --> 00:29:58.652 No one is gonna look out for you, but you. 301 00:29:59.799 --> 00:30:01.019 Okay, a little bit of a tangent there. 302 00:30:02.220 --> 00:30:06.341 I wanted to get back to the futurist lens for this particular story. 303 00:30:07.222 --> 00:30:15.324 Now, the labor cost margin is the metric that is going to reshape a lot of organizational design over the next coming years, maybe even the next decade. 304 00:30:16.405 --> 00:30:23.107 And the fact that this is not a term that most people have even heard of yet does not mean it's not already being applied. 305 00:30:23.707 --> 00:30:26.168 So again, labor cost margin, that's the metric. 306 00:30:27.442 --> 00:30:37.087 And every major company restructuring right now, every efficiency program, every reduction in force, accompanied by a press release about investing in the future, all these things. 307 00:30:38.288 --> 00:30:43.891 There is a labor cost margin as the calculation that goes underneath these decisions. 308 00:30:44.812 --> 00:30:50.755 Now, the reason this matters is because it fundamentally changes the unit of analysis for workforce planning. 309 00:30:51.756 --> 00:30:55.758 Revenue per employee was a productivity measure. 310 00:30:56.665 --> 00:31:00.106 How much output does each human generate? 311 00:31:01.027 --> 00:31:04.568 Labor cost margin is a substitution measure. 312 00:31:05.268 --> 00:31:11.771 How much of what humans used to do can now be done more cheaply by technology? 313 00:31:12.391 --> 00:31:14.332 And what is the optimal mix? 314 00:31:15.092 --> 00:31:16.653 Those are very different questions to be having. 315 00:31:17.273 --> 00:31:22.435 And they lead to very different decisions about hiring, deployment, and organizational structure. 316 00:31:23.215 --> 00:31:34.763 So the early career judgment gap identified in this survey that I've been highlighting for the past 10, 20 minutes here is the most underreported risk in the AI workforce conversation. 317 00:31:35.664 --> 00:31:38.986 Junior roles have historically served two functions, as I've mentioned before. 318 00:31:39.687 --> 00:31:43.169 They produce work and they train the next generation of senior leaders. 319 00:31:43.929 --> 00:31:49.233 And so when you automate the work production function, what else do you do? 320 00:31:49.673 --> 00:31:51.335 You also eliminate the training function. 321 00:31:52.096 --> 00:31:56.002 and a generation of managers who have never had to figure anything out from scratch. 322 00:31:56.747 --> 00:32:02.231 who've always had AI to draft things for them, run their first analysis, build their first model. 323 00:32:02.912 --> 00:32:14.200 This is a generation that can potentially lack any form of deep pattern recognition that comes only through grinding out hard problems without having assistance. 324 00:32:15.121 --> 00:32:17.443 And it's not a problem that shows up in next quarter's earnings. 325 00:32:18.264 --> 00:32:22.467 It's a problem that's gonna show up in five years, in 10 years down the road when those people are running organizations. 326 00:32:24.134 --> 00:32:32.281 The 35% of CEOs spending between 11 and 20% of their entire capital budget on AI is the number I keep coming back to. 327 00:32:33.443 --> 00:32:35.644 It's not experimental at that point. 328 00:32:36.325 --> 00:32:39.206 This is a massive structural commitment. 329 00:32:39.786 --> 00:32:46.089 And at that level of investment, the calculation there around labor cost margin is not gonna be reversed. 330 00:32:46.290 --> 00:32:47.990 I think the direction is set. 331 00:32:48.671 --> 00:33:00.717 The only question is whether organizations are being intentional about what they are building or whether they are simply optimizing for this quarter's cost reductions and calling it transformation. 332 00:33:01.908 --> 00:33:03.489 Story five, last one of the day. 333 00:33:03.690 --> 00:33:15.080 And before I get to that one, quick note, if you're a chief human resource or chief people officer and you find that you're moving beyond the future of work, I'm sorry, beyond traditional HR to focus on the future of work. 334 00:33:16.161 --> 00:33:18.443 And you're tired of expensive HR groups. 335 00:33:18.463 --> 00:33:20.765 You're tired of being upsold and cross sold. 336 00:33:20.785 --> 00:33:24.929 You're tired of fluffy white papers and webinars and all the BS that these groups offer. 337 00:33:26.365 --> 00:33:30.688 then check out futureofworkleaders.com. 338 00:33:31.328 --> 00:33:34.230 I never planned on launching a CHRO group. 339 00:33:34.870 --> 00:33:37.652 I never had intentions of doing this. 340 00:33:37.712 --> 00:33:41.574 I was just speaking and writing books and doing my podcast. 341 00:33:42.254 --> 00:33:46.517 And a lot of CHROs that I talked to said, hey, you know, you talk to a lot of HR executives. 342 00:33:47.237 --> 00:33:52.641 have you ever thought of putting a CHRO group together because we're really interested in focusing on the future of work? 343 00:33:52.741 --> 00:33:54.642 And I thought, all right, I'll give it a shot. 344 00:33:55.243 --> 00:33:56.143 Let's see what happens. 345 00:33:56.663 --> 00:33:57.584 And now here we are. 346 00:33:57.644 --> 00:34:08.652 We have 35 chief human resource officers from some of the world's top companies, including Lego and Tractor Supply and Lumen and Western Digital and lots of amazing, amazing brands out there. 347 00:34:09.132 --> 00:34:10.053 So if you wanna learn more, 348 00:34:11.149 --> 00:34:18.733 and you don't want to be spending a ton of money on these HR groups, go to futureofworkleaders.com. 349 00:34:20.014 --> 00:34:22.795 Okay, getting back into the last story. 350 00:34:23.196 --> 00:34:24.836 Story five, this is from The Verge. 351 00:34:25.037 --> 00:34:29.539 White-collar workers are training the AI that may replace them. 352 00:34:29.679 --> 00:34:38.484 I think I covered this, a version, a headline of this first, I think it was the beginning of the year or maybe end of last year. 353 00:34:39.637 --> 00:34:55.801 So Merkur, you can call it Mercor, Merkur, I've heard it pronounced different ways, M-E-R-C-O-R, a San Francisco-based startup company that connects highly skilled professionals, doctors, lawyers, engineers, investment bankers, consultants, poets, filmmakers, you name it. 354 00:34:56.882 --> 00:35:04.223 They connect these highly specialized and experienced professionals with AI companies that need human expertise to train their models. 355 00:35:04.243 --> 00:35:06.284 And in less than three years, Merkur, you can call it Mercor, Mercor, I've heard it 356 00:35:07.312 --> 00:35:10.954 Merkur has reached a valuation of $10 billion. 357 00:35:11.654 --> 00:35:27.042 They're paying out over $1.5 million daily to its contractors and has signed partnerships with OpenAI and Anthropik, among others, and its CEO is 22 years old. 358 00:35:27.403 --> 00:35:29.344 22 years old. 359 00:35:30.164 --> 00:35:35.447 That's creepy to think about because technically I could be this CEO's dad and... 360 00:35:36.874 --> 00:35:38.074 I still feel like I'm very young. 361 00:35:38.815 --> 00:35:39.455 Apparently I'm not. 362 00:35:40.595 --> 00:35:50.838 Now The Verge went deep on what this company actually does and what it means for the people who are actually out there doing this kind of work. 363 00:35:52.098 --> 00:36:05.222 And the pay ranges from $45 an hour for social media marketing, writing video captions to $150 an hour for Poet, helping AI systems develop poetic structure. 364 00:36:06.190 --> 00:36:08.471 So let me get this straight. 365 00:36:08.511 --> 00:36:13.032 $45 an hour for a social media marketer writing video captions. 366 00:36:13.572 --> 00:36:14.393 A poet. 367 00:36:15.553 --> 00:36:22.655 A poet is making three times the amount as a social media marketer writing video captions. 368 00:36:23.836 --> 00:36:24.156 Amazing. 369 00:36:25.732 --> 00:36:35.960 Um, this goes up to $250 an hour for a dermatologist helping a healthcare partner build clinical decision support tools. 370 00:36:36.461 --> 00:36:40.244 Former Goldman Sachs analysts and McKinsey consultants are among the contractors. 371 00:36:41.024 --> 00:36:43.406 Some moonlight there while keeping their regular jobs. 372 00:36:44.027 --> 00:36:46.469 Many others are doing it out of necessity. 373 00:36:46.509 --> 00:36:52.193 They are between jobs or in a market where the kind of work they built their careers on is getting harder to find. 374 00:36:52.213 --> 00:36:52.273 Um, 375 00:36:53.126 --> 00:36:57.069 The psychological dynamic The Verge captures is worth, I think, thinking about. 376 00:36:58.090 --> 00:37:05.555 One contractor, a video editor named Katie Williams, told the publication, I joked with my friends I'm training AI to take my job someday. 377 00:37:06.395 --> 00:37:16.002 Another contractor, a lawyer named Sarah Kubik, has been supplementing her income through Merkur, but said the experience has actually taught me the limitations of AI. 378 00:37:16.783 --> 00:37:18.844 Some people are going in with clear eyes. 379 00:37:19.545 --> 00:37:22.287 Others are less certain about what they are participating in. 380 00:37:23.007 --> 00:37:31.532 Merker's CEO described the company mission as building a gig economy of expertise, the Uber of high-skilled knowledge work. 381 00:37:32.033 --> 00:37:45.961 The company position is that humans will not run out of meaningful work and that contractors who contribute their knowledge to AI training are actually shaping the outcomes to be more accurate and more thoughtful. 382 00:37:46.902 --> 00:37:50.123 Critics and observers have a harder time with that framing. 383 00:37:50.923 --> 00:37:53.363 Now, the futurist lens here. 384 00:37:55.144 --> 00:37:59.844 This reminds me of, and I'm blinking on what it was called. 385 00:38:00.104 --> 00:38:03.805 So there was, and I'm trying to remember. 386 00:38:04.425 --> 00:38:05.445 Oh, just Googled it. 387 00:38:05.545 --> 00:38:06.626 The Mechanical Turk. 388 00:38:08.326 --> 00:38:10.466 So I don't know if you remember the concept of the Mechanical Turk. 389 00:38:10.846 --> 00:38:13.207 This was an 18th century hoax. 390 00:38:14.078 --> 00:38:17.521 where a machine appeared to be playing chess. 391 00:38:18.201 --> 00:38:32.091 So this was a box, and people would go up, and you would make a move on the chessboard, and seemingly the box, the chessboard, would make moves as the opponent. 392 00:38:32.431 --> 00:38:33.412 And again, 18th century. 393 00:38:34.933 --> 00:38:39.196 But what was later discovered is that inside this box, there was a human character, 394 00:38:40.741 --> 00:38:43.483 underneath the chessboard, making the moves. 395 00:38:44.523 --> 00:38:53.889 And so it wasn't actually a mechanical Turk, a mechanical automaton, or a mechanical robot making the moves, as so many people thought. 396 00:38:54.169 --> 00:38:58.731 It was actually a human being underneath the chessboard, inside this giant box, making the moves. 397 00:39:00.512 --> 00:39:03.074 And so when I hear these stories, it's very interesting. 398 00:39:04.215 --> 00:39:08.197 On the one hand, you have platforms like OpenAI, 399 00:39:09.507 --> 00:39:13.729 who already have an integration to analyze your health records. 400 00:39:14.409 --> 00:39:25.593 So right now you can go to OpenAI, you can go into your chat GPT, there is the health option, and you can sync it with your UCLA medical records, your Kaiser medical records, whatever you want. 401 00:39:26.734 --> 00:39:33.676 And it will give you an analysis of your medical records, you can ask it for recommendations, you can do all the things that you would if you're going to a doctor. 402 00:39:35.797 --> 00:39:38.278 So you see these types of stories, and on the one hand you're like, wait a minute, 403 00:39:39.823 --> 00:39:46.568 On the one hand, a dermatologist is getting paid $250 to train supposedly this AI. 404 00:39:47.889 --> 00:39:56.134 But then you're also telling me that AI can already give me a medical analysis and look at my medical records and give me suggestions. 405 00:39:56.175 --> 00:39:57.175 So it's kind of like, which one is it? 406 00:39:58.216 --> 00:40:05.081 Is AI advanced so much that it can do all these amazing things and diagnose you and automate jobs or... 407 00:40:06.544 --> 00:40:10.308 Are we still having a social media marketing person teach AI how to write video descriptions? 408 00:40:10.789 --> 00:40:11.650 It's kind of like, wait a minute. 409 00:40:12.371 --> 00:40:14.233 The narrative doesn't make sense here, right? 410 00:40:14.353 --> 00:40:17.737 And hopefully you can see why it doesn't make sense. 411 00:40:18.698 --> 00:40:23.183 I mean, right now, AI can already write a lot of captions. 412 00:40:23.203 --> 00:40:24.865 It can do all of these things. 413 00:40:25.957 --> 00:40:29.100 So the narrative seems to be pretty inconsistent. 414 00:40:29.440 --> 00:40:33.744 Either we're still teaching these AI tools or AI can already do this stuff. 415 00:40:33.764 --> 00:40:41.430 And I'm actually on the Merkur website right now and I'm just scrolling through looking at some of the options here. 416 00:40:41.470 --> 00:40:44.252 So if you're a math PhD, $73 an hour. 417 00:40:45.013 --> 00:40:47.355 Physics PhD, $73 an hour. 418 00:40:48.856 --> 00:40:53.120 Let's see, who are some of the higher paid jobs here that we can see? 419 00:40:53.300 --> 00:40:54.081 Most pay. 420 00:40:55.075 --> 00:41:00.917 Okay, let's do a search for most pay and see what these roles are. 421 00:41:02.418 --> 00:41:06.639 Okay, so legal content reviewer, 140 bucks an hour. 422 00:41:06.659 --> 00:41:11.621 It doesn't say for some of them, some of them are hidden on here as well. 423 00:41:12.301 --> 00:41:13.342 But you know, 120, $130 an hour. 424 00:41:13.362 --> 00:41:13.682 Oh my goodness. 425 00:41:13.702 --> 00:41:15.522 Hebrew professional voice actor, 45 to 225 an hour. 426 00:41:24.254 --> 00:41:42.064 um transactional regulatory associate up to 200 an hour japanese professional voice actor 200 an hour so if you can do a hebrew professional voice and a japanese professional voice you can kind of double dip there a little bit 427 00:41:43.787 --> 00:41:50.890 So very, very interesting expert in equities research, a law librarian, Hindi professional actor. 428 00:41:50.910 --> 00:41:53.611 So lots of different roles that are on this site. 429 00:41:54.511 --> 00:42:01.653 Now, I get the acting and the voice is one thing, but this to me seems a very, you know, somewhat paradoxical, so to speak. 430 00:42:02.814 --> 00:42:06.515 Now, I think this is a pretty interesting topic. 431 00:42:08.030 --> 00:42:15.801 trend to be paying attention to because Merkur has not just built a business model, it's a mirror held up to the entire AI transition period that we're thinking through. 432 00:42:16.701 --> 00:42:27.590 So we have a generation of highly educated, highly skilled professionals who built their careers on expertise that is now being systematically encoded into machines. 433 00:42:28.531 --> 00:42:37.759 Some of them are actually doing it voluntarily, voluntarily, at good hourly rates with a sense of agency about shaping the AI that will work alongside them. 434 00:42:38.580 --> 00:42:40.722 Others are doing it because the job market is, you know, crazy. 435 00:42:43.387 --> 00:42:45.828 let's just say, messed up for their traditional roles. 436 00:42:46.449 --> 00:42:48.250 It has tightened and they need the income now. 437 00:42:49.050 --> 00:42:55.234 And the machine they are training will over time make it harder for others just like them to find the kind of work they used to do. 438 00:42:56.094 --> 00:43:02.478 And there is a name for this dynamic in economic history and the name is the transitional trap. 439 00:43:03.278 --> 00:43:13.882 Workers who are displaced by a new technology often becomes the most efficient means of accelerating its adoption because they have the domain expertise and they need the money. 440 00:43:14.882 --> 00:43:28.487 So the workers, again, who are displaced by a new technology often become the most efficient means of accelerating its adoption because they have the domain expertise and they need the money. 441 00:43:29.128 --> 00:43:29.808 Coal miners... 442 00:43:30.678 --> 00:43:45.668 who became diesel mechanics, typographers, who taught desktop publishing, and now doctors, lawyers, investment bankers, training the AI models that will automate the entry-level and mid-level versions of their professions. 443 00:43:46.448 --> 00:43:48.990 The pattern repeats, but the scale this time is different. 444 00:43:49.817 --> 00:43:53.921 Now that Merkur benchmark data is also, I think, worth paying attention to here. 445 00:43:54.621 --> 00:44:07.872 Merkur's own research, the Apex Agents Benchmark, found that the best AI agents currently complete fewer than 25% of real white-collar professional tasks in one shot. 446 00:44:08.493 --> 00:44:13.938 Even with multiple attempts, no model is ready to replace a professional yet end-to-end. 447 00:44:14.998 --> 00:44:24.767 That means we're in a window where humans are still essential, but actively training the systems will eventually reduce that essentialness. 448 00:44:25.668 --> 00:44:31.093 The window is open, but perhaps it may not stay open forever. 449 00:44:31.113 --> 00:44:32.334 It may not stay open indefinitely. 450 00:44:32.374 --> 00:44:33.255 So for business leaders... 451 00:44:34.097 --> 00:44:39.420 The Merkur story surfaces a question that most organizations have not formally asked themselves. 452 00:44:40.220 --> 00:44:52.026 What happens to the institutional knowledge when the people who built it start exporting it into AI systems, either through their employers, internal tools, or through platforms just like Merkur? 453 00:44:52.970 --> 00:44:59.472 It's not a hypothetical risk, it's actually happening right now, at scale and in every single industry. 454 00:45:00.172 --> 00:45:04.953 And the organizations that I think treat this as a future problem are already one step behind. 455 00:45:05.653 --> 00:45:12.355 The lawyer who said training AI taught her its limitations might be the most important voice in this entire story. 456 00:45:12.975 --> 00:45:21.818 Because understanding what technology cannot do, not just what it can do, is the human capability that is going to matter most in the decade ahead. 457 00:45:22.558 --> 00:45:25.520 And it's not a skill that comes from observing AI from a distance. 458 00:45:25.560 --> 00:45:31.703 It comes from working closely with it, interacting with it, challenging it, using it, finding the edges. 459 00:45:32.563 --> 00:45:45.490 And the people doing that work right now, even if they're doing it just as a side gig, to pay their bills are building exactly the kind of judgment that will be scarce and valuable as this transition continues. 460 00:45:46.390 --> 00:45:48.491 So those are the top stories of the day. 461 00:45:48.531 --> 00:45:51.953 And I want to just leave you with one thing to think about. 462 00:45:52.535 --> 00:45:58.138 Just one little breadcrumb in there, something for you to gnaw on. 463 00:45:59.459 --> 00:46:02.020 And that is, what happens if nothing changes? 464 00:46:03.040 --> 00:46:04.881 What happens if this is all just a bunch of BS? 465 00:46:05.702 --> 00:46:08.603 What happens if in five years you're still listening to this podcast? 466 00:46:10.084 --> 00:46:13.926 God willing, I'm still doing this podcast, and we're still having this conversation. 467 00:46:15.242 --> 00:46:17.624 Agents have not yet made their way into the workplace. 468 00:46:17.704 --> 00:46:19.705 Companies still not seeing ROI. 469 00:46:20.246 --> 00:46:25.350 Next five years from 2035 to 2040 is going to be so impactful, blah, blah, blah, blah, blah. 470 00:46:27.411 --> 00:46:29.152 How come we don't have those conversations? 471 00:46:30.353 --> 00:46:32.315 Because that is one of the potential futures. 472 00:46:33.356 --> 00:46:39.440 Is that maybe, just maybe, things will kind of stay as they are now. 473 00:46:40.261 --> 00:46:42.803 A little bit of disruption, a little bit of change. 474 00:46:42.823 --> 00:46:44.204 You know, that's common. 475 00:46:44.224 --> 00:46:44.304 But, 476 00:46:45.175 --> 00:46:49.018 But maybe things won't be that different than they are now. 477 00:46:49.418 --> 00:46:52.520 Again, one potential future to be thinking about. 478 00:46:52.841 --> 00:47:06.630 We're so obsessed with change and transformation and evolution and displacement and augmentation that maybe, just maybe, we'll be doing the same freaking stuff we're doing now in five years. 479 00:47:07.571 --> 00:47:11.394 Again, I'm not saying that is the only future that will happen, but it's one potential. 480 00:47:11.954 --> 00:47:12.635 So I'm going to leave you... 481 00:47:13.549 --> 00:47:15.031 with that little nugget to think about. 482 00:47:16.612 --> 00:47:21.397 If you get a couple seconds, please rate the podcast on Apple Podcasts or on Spotify. 483 00:47:22.098 --> 00:47:36.492 I keep looking at our analytics here and most of the people who listen to this podcast, and there are thousands of you on Spotify, on Apple, on YouTube, you listen to it and I appreciate it very much, but most of you are not actually subscribed to the show. 484 00:47:37.555 --> 00:47:39.736 And the subscription numbers really help. 485 00:47:39.956 --> 00:47:44.998 It helps the show get discovered on platforms like Apple and Spotify and YouTube. 486 00:47:45.559 --> 00:47:51.001 So if you get a couple of seconds, please do me a favor, hit that subscribe or follow button and leave a rating. 487 00:47:51.321 --> 00:47:54.903 The ratings really make a difference on Spotify and on Apple. 488 00:47:55.383 --> 00:47:58.064 Help cut through the algorithms, help other people discover the show. 489 00:47:58.604 --> 00:48:04.567 And it helps me and my team here bring in more amazing guests and create more episodes just like this. 490 00:48:05.287 --> 00:48:06.429 So thank you so much for tuning in. 491 00:48:06.469 --> 00:48:14.700 And my email, if you're interested in sponsoring the program or having me speak or just have feedback from me, is jacob at thefutureorganization.com. 492 00:48:14.740 --> 00:48:18.726 Again, that's jacob at thefutureorganization.com. 493 00:48:19.267 --> 00:48:20.388 Have a wonderful rest of the day. 494 00:48:20.869 --> 00:48:21.570 I'll be back tomorrow.